Float Gallery

Interactive visualization of various floating point formats.

You can click on the bits to toggle them or use the input box to enter a floating point number

Float32

Good old IEEE 754 32-bit floating point number.

TensorFloat32

More info on TensorFloat-32

Float16

IEEE 754 half percision floating point number.

BFloat16

BFloat16 has larger range than float16 but less precision. Invented by Google Brain.

OCP FP8 E4M3

OCP FP8 E4M3 is a compact floating point format with 4-bit exponent and 3-bit mantissa. Doesn't support infinity. OCP FP8 Specification.

OCP FP8 E5M2

OCP FP8 E5M2 is a compact floating point format with 5-bit exponent and 2-bit mantissa. OCP FP8 Specification.

FP8 E4M3 FNUZ

FP8 E4M3 FNUZ is a floating point format with 4-bit exponent and 3-bit mantissa. It has a special encoding for NaN and doesn't support infinity.

FP8 E5M2 FNUZ

FP8 E5M2 FNUZ is a floating point format with 5-bit exponent and 2-bit mantissa. It has a special encoding for NaN and doesn't support infinity.

OCP FP6 E2M3

OCP FP6 E2M3 is a floating point format with 2-bit exponent and 3-bit mantissa. Doesn't support NaN or infinity. OCP Microscaling Formats Specification

OCP FP6 E3M2

OCP FP6 E3M2 is a floating point format with 3-bit exponent and 2-bit mantissa. Doesn't support NaN or infinity. OCP Microscaling Formats Specification

OCP FP4

OCP FP4 is a floating point format with 2-bit exponent and 1-bit mantissa. Doesn't support NaN or infinity. OCP Microscaling Formats Specification